Friday, February 24, 2012

LIES, DAMNED LIES, AND STATISTICS (take 2)

"Science will beat pseudo-science every time." --Christopher Hitchens

Ignorance is (not necessarily) bliss. And spreading misinformation and drawing false conclusions on poorly conducted studies is shameful. Even if you are attacking something that seems intrinsically bad--like soda (everyone likes a scapegoat). I was going to post the following remark on a shared link on facebook, but I didn't want to stir the pot with this particular person (she's an english major). Instead I will post it here so I can get it off my unscathed chest:

This is nothing more than scare tactics to get people to click on their site. The last line says it all. "However, people that drink diet soda tend to have more unhealthy lifestyles, and so the study can't specifically be linked to diet soda as a cause." In other words, correlation means nothing by itself (statistics 101). It can only support other evidence. I have seen similar "studies" attacking meat and high fructose corn syrup using the same tactics. All of these things may very well be bad for you, but these sorts of studies only hurt their credibility by misrepresenting the data. As Mark Twain said, "There are three kinds of lies: lies, damned lies and statistics."